Posted: Wed Oct 31, 2007 9:05 pm
Wikipedia is an interesting example - it is recognized there that items may be public domain in some countries but not others. To quote from their public domain page (http://en.wikipedia.org/wiki/Wikipedia:Public_domain) " "there is no such thing as the public domain on the Internet. " It's a very instructive page.
It's worth bearing in mind that their servers are in Florida so they take special care to follow U.S. copyright law and also pay attention to the Berne Convention. In cases where something is P.D. in one country but not in others (especially within the Berne Convention) then generally it is not considered "sufficiently public domain" for Wikipedia. An interesting case is the pre-1923 rule. Anything published anywhere in the world before 1923, regardless of when the creator died, is P.D. in the U.S. but may well be copyright elsewhere under "life + 50 years" or "life + 70 years" rules. If Wikipedia was in the business of "maximising the availability of public domain material" (i.e. "getting away with as much as possible") the community there might choose to allow any pre-1923 publication, since it would be out of copyright within the U.S. where the Wikipedia servers are located. However, Wikipedia policy is not to try to get away with as much as possible, rather it restricts the use of the pre-1923 rule to publications published within the U.S. and generally enforces "life + 70" to publications outside the U.S. with a few exceptions. "Life + 50" suffices for Canada, but this is only used for works that originate in Canada, not were merely republished there. The basic idea is for the work to be P.D. both in the U.S. where the servers are, and also in the place where the work originated.
As has already been discussed, the situation with sheet music is complicated because the original composer, any subsequent arranger, the setting for a publication and possibly even the editing for certain editions may all have intellectual property implications. But IMSLP had an excellent system for copyright monitoring. The hard work has already been done sorting through the music and putting on appropriate copyright tags.
The strength of that system is that now it would not be technically impossible to implement either of the two obvious solutions - the universal application of the "strictest" copyright rules would be easy or the finer approach country-by-country IP block (identifying which countries something is allowed would not be hard to work out from the copyright tags - admittedly you might not know the law of Mali, but unless Malian lawyers write to you, I wouldn't worry about it; the major internet audiences of the US, EU, Australia/NZ, Canada, Japan and South Korea are all relatively easy cases).
Now, I for one hate the "life + 70" term, I think it's ridiculously restrictive. But you can see there's a problem if a work composed by a German, published first in Germany, and still copyright in Germany, can be accessed by a German in Germany using the IMSLP - it doesn't have to printed, the mere appearance of the work on the screen is technically an instance of "publication". At this point you can understand why the copyright holder might start to take a (legitimate) interest. (Not saying that I approve in any way of the heavy-handed approach they took, but it's clear that IMSLP was leaving itself open to legal scrutiny.)
Even Wikipedia doesn't try to get away with all it could. Wikipedia's copyright policies are self-regulated, and many disputes arise over low quality reproductions of works created long ago by obscure photographers/artists/designers/musicians and which are pretty clearly of no commercial value, with scant chance of being reprinted. IMSLP is treading on hotter coals already - even the works it publishes (and legally this is what it's doing) high quality digital reproductions of and which are P.D. worldwide are often being reprinted for profit by other organisations; those works only P.D. in certain jurisdictions may not be the hottest of properties but they do still make money for their copyright holders. Basically the "get away with as much as possible" model is bound to get the legal eagles involved. Bearing this in mind, a more conservative approach to declaring work "public domain" seems sensible.
Wikipedia might provide an example of a "third way", less draconian than imposing the strictest possible interpretation of P.D. but less technically difficult than the case-wise IP block. What about trying to respect the law for where the material is sourced from, and the law where the servers are? The work, first published in Canada/South Korea, of a Canadian/South Korean composer could benefit from the more relaxed "life + 50" rule, if it's a German composer published first in Germany then play it safe with "life + 70". It might not be 100% legally watertight - perhaps the holder to the rights in Germany of the South Korean composer will kick up a fuss - but such problems seem unlikely (and can always be responded to by pulling the appropriate content) and it would be a lot determining the copyright status in one or two countries rather than all 190-ish. In a sense, it's unfortunate that IMSLP happened to be based in Canada and therefore tempted fate more by applying Canada's unusually "liberal" copyright laws (I think life+50 is still pretty tight really, but unfortunately it compares well internationally) - had the site been created by an American or German then odds are this problem wouldn't have arisen because those harsher laws would have seemed more natural and almost certainly been complied with. (This is in no way a personal criticism, just a reflection of the laws of chance - the U.S. has 301 million people, the E.U. 490 million, and Canada only 33 million, but in this case the 4% chance of being Canadian actually occurred!)
Any added conservatism on declaring works "sufficiently" P.D. will result in the loss of some material (although if the IP blocking system was implemented, it'd only apply to appropriate users and not to the central library itself). It's worth bearing in mind that the copyright laws that U-E wanted applied probably affect the vast majority of users (not just a small minority) but only the minority of IMSLP material. Unlike some other posters here, I don't think that preventing users from downloading works illegally (in their jurisdiction at least) would either devastate functionality or make all U.S. users incredibly unhappy, unless the font of happiness and productivity is the performance of illegal acts. Moreover, that attitude significantly undervalues the amazing work done at IMSLP, in the uploading, cataloguing, and careful copyright identifying of thousands of (often rare) manuscripts which are undoubtedly copyright-free worldwide. That work is incredibly valuable and easily outweighs the value of the copyright-disputed material. There's no need to feel guilty about the copyrighted material, since it was uploaded in good faith and with a careful eye on Canadian law, but whether it's worth a legal wrangle over isn't a question for me to judge. But I do think it's worth pointing out (1) that there are lower-risk strategies for publishing material on the web that would substantially reduce the chances of legal problems, and (2) that the disputed material only ever happened to be on here in the first place due to the less than 5% chance of the creator being Canadian. I don't want to sound like I'm either siding with the bullies, or sticking up for absurd copyright terms, but I do think there's so much incredibly valuable work at IMSLP that is absolutely unaffected by modern copyright law that it would be a great shame if it was lost due to some dumb legal dispute...
As a previous "happy customer" of this site I would like to thank all who've worked on it, particularly the creator, and I really hope that some solution can be found!
It's worth bearing in mind that their servers are in Florida so they take special care to follow U.S. copyright law and also pay attention to the Berne Convention. In cases where something is P.D. in one country but not in others (especially within the Berne Convention) then generally it is not considered "sufficiently public domain" for Wikipedia. An interesting case is the pre-1923 rule. Anything published anywhere in the world before 1923, regardless of when the creator died, is P.D. in the U.S. but may well be copyright elsewhere under "life + 50 years" or "life + 70 years" rules. If Wikipedia was in the business of "maximising the availability of public domain material" (i.e. "getting away with as much as possible") the community there might choose to allow any pre-1923 publication, since it would be out of copyright within the U.S. where the Wikipedia servers are located. However, Wikipedia policy is not to try to get away with as much as possible, rather it restricts the use of the pre-1923 rule to publications published within the U.S. and generally enforces "life + 70" to publications outside the U.S. with a few exceptions. "Life + 50" suffices for Canada, but this is only used for works that originate in Canada, not were merely republished there. The basic idea is for the work to be P.D. both in the U.S. where the servers are, and also in the place where the work originated.
As has already been discussed, the situation with sheet music is complicated because the original composer, any subsequent arranger, the setting for a publication and possibly even the editing for certain editions may all have intellectual property implications. But IMSLP had an excellent system for copyright monitoring. The hard work has already been done sorting through the music and putting on appropriate copyright tags.
The strength of that system is that now it would not be technically impossible to implement either of the two obvious solutions - the universal application of the "strictest" copyright rules would be easy or the finer approach country-by-country IP block (identifying which countries something is allowed would not be hard to work out from the copyright tags - admittedly you might not know the law of Mali, but unless Malian lawyers write to you, I wouldn't worry about it; the major internet audiences of the US, EU, Australia/NZ, Canada, Japan and South Korea are all relatively easy cases).
Now, I for one hate the "life + 70" term, I think it's ridiculously restrictive. But you can see there's a problem if a work composed by a German, published first in Germany, and still copyright in Germany, can be accessed by a German in Germany using the IMSLP - it doesn't have to printed, the mere appearance of the work on the screen is technically an instance of "publication". At this point you can understand why the copyright holder might start to take a (legitimate) interest. (Not saying that I approve in any way of the heavy-handed approach they took, but it's clear that IMSLP was leaving itself open to legal scrutiny.)
Even Wikipedia doesn't try to get away with all it could. Wikipedia's copyright policies are self-regulated, and many disputes arise over low quality reproductions of works created long ago by obscure photographers/artists/designers/musicians and which are pretty clearly of no commercial value, with scant chance of being reprinted. IMSLP is treading on hotter coals already - even the works it publishes (and legally this is what it's doing) high quality digital reproductions of and which are P.D. worldwide are often being reprinted for profit by other organisations; those works only P.D. in certain jurisdictions may not be the hottest of properties but they do still make money for their copyright holders. Basically the "get away with as much as possible" model is bound to get the legal eagles involved. Bearing this in mind, a more conservative approach to declaring work "public domain" seems sensible.
Wikipedia might provide an example of a "third way", less draconian than imposing the strictest possible interpretation of P.D. but less technically difficult than the case-wise IP block. What about trying to respect the law for where the material is sourced from, and the law where the servers are? The work, first published in Canada/South Korea, of a Canadian/South Korean composer could benefit from the more relaxed "life + 50" rule, if it's a German composer published first in Germany then play it safe with "life + 70". It might not be 100% legally watertight - perhaps the holder to the rights in Germany of the South Korean composer will kick up a fuss - but such problems seem unlikely (and can always be responded to by pulling the appropriate content) and it would be a lot determining the copyright status in one or two countries rather than all 190-ish. In a sense, it's unfortunate that IMSLP happened to be based in Canada and therefore tempted fate more by applying Canada's unusually "liberal" copyright laws (I think life+50 is still pretty tight really, but unfortunately it compares well internationally) - had the site been created by an American or German then odds are this problem wouldn't have arisen because those harsher laws would have seemed more natural and almost certainly been complied with. (This is in no way a personal criticism, just a reflection of the laws of chance - the U.S. has 301 million people, the E.U. 490 million, and Canada only 33 million, but in this case the 4% chance of being Canadian actually occurred!)
Any added conservatism on declaring works "sufficiently" P.D. will result in the loss of some material (although if the IP blocking system was implemented, it'd only apply to appropriate users and not to the central library itself). It's worth bearing in mind that the copyright laws that U-E wanted applied probably affect the vast majority of users (not just a small minority) but only the minority of IMSLP material. Unlike some other posters here, I don't think that preventing users from downloading works illegally (in their jurisdiction at least) would either devastate functionality or make all U.S. users incredibly unhappy, unless the font of happiness and productivity is the performance of illegal acts. Moreover, that attitude significantly undervalues the amazing work done at IMSLP, in the uploading, cataloguing, and careful copyright identifying of thousands of (often rare) manuscripts which are undoubtedly copyright-free worldwide. That work is incredibly valuable and easily outweighs the value of the copyright-disputed material. There's no need to feel guilty about the copyrighted material, since it was uploaded in good faith and with a careful eye on Canadian law, but whether it's worth a legal wrangle over isn't a question for me to judge. But I do think it's worth pointing out (1) that there are lower-risk strategies for publishing material on the web that would substantially reduce the chances of legal problems, and (2) that the disputed material only ever happened to be on here in the first place due to the less than 5% chance of the creator being Canadian. I don't want to sound like I'm either siding with the bullies, or sticking up for absurd copyright terms, but I do think there's so much incredibly valuable work at IMSLP that is absolutely unaffected by modern copyright law that it would be a great shame if it was lost due to some dumb legal dispute...
As a previous "happy customer" of this site I would like to thank all who've worked on it, particularly the creator, and I really hope that some solution can be found!